CN105682231A - Method for joint distribution of power and time for cooperative communication of cognitive radio network - Google Patents

Abstract

The invention discloses a method for joint distribution of power and time for cooperative communication of a cognitive radio network. Master users in the cognitive radio network are taken as relay to cooperate with slave users to complete transmission, joint distribution is carried out on the power and the time, and when the optimal power is selected to distribute, the optimal distribution of the time is considered. The original distribution problem is decomposed into two sub-problems of outer layer power distribution and inner layer power distribution, and an auction method is utilized to carry out inner layer transmission time distribution on the slave users based on the given transmission power of the master users; and a local search method is utilized to distribute the transmission power of the master users at the outer layer. According to the method, the spectral efficiency is effectively improved and the total utility of the master users is maximized. On the one hand, each master user capable of maximizing the utility of the corresponding slave user is searched for each slave user, and the cooperative transmission is carried out in the distributive time; and on the other hand, the requirements for limitation of the total power between the master users are met.

Description

The power of cognitive radio networks collaboration communication and temporal joint distribution method

Technical field

The present invention relates to the power of a kind of network cooperation communication and temporal joint distribution method, the power of especially a kind of cognitive radio networks collaboration communication and temporal joint distribution method, belong to field of resource allocation in cordless communication network.

Technical background

Radio Spectrum Resource is a kind of limited natural resources, is generally used by government authorization. Along with developing rapidly of wireless communication technology, frequency spectrum resource becomes more and more nervous, but many investigation find that a large amount of mandate frequency spectrum is idle or utilization rate is extremely low. WLAN (WLAN) technology, wireless personal area network (WPAN) technology, wireless MAN (WMAN) technology develop rapidly in recent years, and increasing radio communication service techniques described above is linked into the Internet. These network technologies are used mostly unauthorized frequency range (UFB) work. Along with people are increasing to the demand of unauthorized frequency range frequency spectrum, the problem that frequency spectrum resource is deficient is day by day serious. And current static frequency spectrum distribution system causes resource average utilization low and can not provide extra utilizable bandwidth for the radio communication service of high speed data transfers.

For this irrational frequency spectrum usage mode, it is necessary to the frequency spectrum resource that those are underutilized by introducing dynamic spectrum access technology is used, it is achieved the dynamically distribution of frequency spectrum is thus being greatly enhanced the utilization rate of frequency spectrum. Cognitive radio (CognitiveRadio, CR) be in order to alleviate growing radio spectrum resources demand and frequency spectrum resource in short supply between contradiction and a kind of new technique of proposing. Cognitive radio technology is by allowing from user (SecondaryUser, SU) perceived spectral resource adaptively, take idle frequency spectrum to opportunistic, as long as not to primary user (PrimaryUser, PU) produce under the premise of harmful interference, primary user's frequency range can be accessed from user, improve the availability of frequency spectrum, the problem solving the waste of present stage frequency spectrum resource to the full extent.

Cognitive radio networks utilizes cooperative communication technology, it is possible to while improving transmission performance, improve frequency spectrum service efficiency as much as possible. Collaboration communication in existing cognition network, mostly considers to assist primary user to transmit from user, from user's relaying as primary user, to compensate the loss using frequency spectrum that primary user is caused from user.

Network resource allocation method in cognitive radio system mainly has two kinds, and one is centralized, and one is distributed.Centralized resource allocation methods needs to complete the calculating of complexity at server end, and the designing requirement of network model is higher. Complicated calculating is then distributed at each user side by distributed resource allocation methods, and in existing distributed resource allocation method, game theory is considered as an effective instrument, can not only improve the utilization rate of resource, and can effectively reduce computational complexity.

Summary of the invention

Present invention aims to the defect that prior art exists, it is provided that the power of a kind of cognitive radio networks collaboration communication and temporal joint distribution method.

For achieving the above object, the present invention adopts following technical proposals:

The power of a kind of cognitive radio networks collaboration communication and temporal joint distribution method, described cognitive radio networks includes base station PBS, a M primary user PU_m, m=1LM and N number of from user SU_n, n=1LN; M primary user communicates with base station, forms uplink network; Primary user PU_mBusy channel m, its through-put power is P_m; The acceptable power of base station PBS is P_tot, the vector power P=(P of primary user₁,L,P_M) determined by base station PBS, and meetFrom user SU_nTransmitting terminal be ST_n, receiving terminal is SR_n; Primary user channel is distributed to need collaboration communication from user, each channel can only be assigned at most one within the same time and use from user, each can only assign at most a channel from user; Primary user PU_mThe overall transmission time that can be used to distribute is T_m, primary user PU_mUnit interval price be c_m, from user SU_nObtain the primary user PU of its cooperation transmission_mThe time distributed is t_mn; Primary user PU_mUtility function u_mRepresent, and definition $u_m=\underset{n}{\Σ}{t}_{mn}\·c_m;$

Comprise the following steps:

Step 1: base station PBS initialization power vector P=(P₁,L,P_M), and add in candidate power vector set U;

Step 2: initial internal layer time distribution: setting up distribution model of initial internal layer time and solve and obtain primary user's overall utility, it can be used as the initial value of primary user's overall utility maximum, distribution of described initial internal layer time model is:

$\underset{t}{\max mize}\underset{m}{\Σ}{u}_m=\underset{m}{\Σ}\underset{n}{\Σ}{t}_{mn}\·c_m---\left(1\right)$

It meets constraints:

$\underset{n}{\Σ}{t}_{mn}\≤T_m,\∀m---\left(2\right)$

Step 3: take out a candidate power vector from described candidate power vector set as current candidate vector power;

Step 4: outer layer power distributes: generate all neighbours' vector powers of described current candidate vector power;

Step 5: each neighbours' vector power of described current candidate vector power carries out neighbours' internal layer time to be distributed: set up neighbours' internal layer time and distribute model and solve and obtain corresponding primary user's overall utility, updating described candidate power vector set and primary user's overall utility maximum, wherein neighbours' internal layer time distribution model is:

$\underset{t}{\max mize}\underset{m}{\Σ}{u}_m=\underset{m}{\Σ}\underset{n}{\Σ}{t}_{mn}\·c_m---\left(3\right)$

It meets constraints:

$\underset{n}{\Σ}{t}_{mn}\≤T_m,\∀m---\left(4\right)$

Step 6: judge whether described candidate power vector set is empty, if it is, turn to step 7; Otherwise, step 3 is turned to;

Step 7: complete power and time distribution, start cooperation transmission: power and time when reaching maximum according to making primary user's overall utility distribute, all primary users distribution is from the request time of user, and during this period of time assists to communicate from user with corresponding power.

Base station PBS initialization power vector P=(P in described step 1₁,L,P_M), and the method adding candidate power vector set U, it is made up of step in detail below:

Step 1-1: set up candidate power vector set U;

Step 1-2: by initialization power vector P=(P₁,L,P_M) add in described candidate power vector set U.

Described step 2 solves the method for distribution of initial internal layer time model, is made up of step in detail below:

Step 2-1: according to described initialization power vector, each primary user PU_mCarry out parameter initialization, and be broadcasted to all of from user; Described carry out initialized parameter include send power of information P_m, timing parameter τ, price step-length δ, reservation priceTiming parameter τ initial value is set to 0;

Step 2-2: from user SU_nCalculate relative to each primary user PU_mTransmission rate R_mn:

$R_{mn}=W\·log(1+{\mathrm{\Γ}}_{mn}(1)+{\mathrm{\Γ}}_{mn}(2\left)\right);{\mathrm{\Γ}}_{mn}\left(1\right)=\frac{Q_n\·h_{ss}^{nm}}{N_0};{\mathrm{\Γ}}_{mn}\left(2\right)=\frac{P_m\·h_{ps}^{nm}}{N_0}---\left(5\right)$

Wherein, Γ_mnAnd Γ (1)_mn(2) represent respectively from user SU_nTransmitting terminal ST_nWith primary user PU_mIt send information to from user SU_nReceiving terminal SR_nSignal to noise ratio,WithRepresent respectively from user SU_nTransmitting terminal ST_nArrive from user SU_nReceiving terminal SR_nWith primary user PU_mArrive from user SU_nReceiving terminal SR_nChannel gain, W represents bandwidth, Q_nRepresent from user SU_nThe power of transmission information, P_mRepresent primary user PU_mThe power of forwarding information, N₀Represent noise variance;

Step 2-3: from user SU_nCalculate relative to each primary user PU_mValue of utility u_mn:

$u_{mn}={(t_{mn}\·R_{mn})}^b-c_m^{\mathrm{\τ}}\·t_{mn}---\left(6\right)$

Wherein b is the satisfaction factor from user, represents the speed impact on value of utility and 0 < b < 1;

Step 2-4: from user SU_nCalculate relative to each primary user PU_mOptimal timeWith corresponding value of utility

$\frac{\&part;u_{mn}}{\&part;t_{mn}}=0\&DoubleRightArrow;t_{mn}\left(c_m^{\mathrm{\&tau;}}\right)={(b\&CenterDot;{R_{mn}}^b)}^{1/(1-b)}\&CenterDot;{\left(c_m^{\mathrm{\&tau;}}\right)}^{1/(b-1)}---\left(7\right)$

$u_{mn}\left(c_m^{\mathrm{\&tau;}}\right)=(b^{b/(1-b)}-b^{1/(1-b)})\&CenterDot;{R_{mn}}^{b/(1-b)}\&CenterDot;{\left(c_m^{\mathrm{\&tau;}}\right)}^{b/(b-1)}---\left(8\right)$

Step 2-5: from user SU_nDetermining the primary user making its value of utility maximum from M primary user, its numbering is designated as m_n:

$m_n=\arg \underset{x}{max}{u}_{mn}\left(c_m^{\mathrm{\&tau;}}\right)---\left(9\right)$

Step 2-6: from user SU_nDetermine and primary userCorresponding optimum request timeAnd optimum request timeFeed back to primary user

Step 2-7: each primary user PU_mThe all optimum request times from user's request received are sued for peace:

$T_m^{\mathrm{\&tau;}}\left(c_m^{\mathrm{\&tau;}}\right)=\underset{n=1}{\overset{N}{\&Sigma;}}{t}_{mn}\left(c_m^{\mathrm{\&tau;}}\right)---\left(10\right)$

Step 2-8: judge each primary user PU_mWhether all satisfiedIf it is, turn to step 2-10; Otherwise, step 2-9 is turned to;

Step 2-9: timing parameter changes into τ=τ+1; All satisfiedPrimary user PU_mRetention time price is constant, namelyAll it is unsatisfactory forPrimary user PU_mRenewal time priceAnd by the time price c after renewal_mIt is published to N number of from user; Turn to step 2-3;

Step 2-10: timing parameter during note End of Auction is τ=L, according to each from user SU_nRequest timeCalculate primary user's overall utility, and as primary user overall utility maximum u_maxInitial value:

$u_{max}=\underset{m}{\&Sigma;}{u}_m=\underset{m}{\&Sigma;}\underset{n}{\&Sigma;}{t}_{mn}^L\&CenterDot;c_m^L---\left(11\right)$

Outer layer power distribution in described step 4, the method solving all neighbours' vector powers of current candidate vector power, it is made up of step in detail below:

Step 4-1: definition unit transfer power, is designated as Δ P, and meets Δ P < < P_tot/ M;

Step 4-2: in described current candidate vector power, arbitrarily select two primary users, the power of one of them primary user is reduced or increases unit transfer power Δ P, and the power of another primary user correspondingly increases or reduces unit transfer power Δ P, satisfy condition in the process the power of the power of all primary users and constant and each primary user on the occasion of, namelyP_m> 0; Travel through all situations meeting above-mentioned power transfer process, obtain all neighbours' vector powers of described current candidate vector power.

The method for solving that described neighbours' internal layer time distributes in described step 5 model and the process updating described candidate power vector set and primary user's overall utility maximum are made up of step in detail below:

Step 5-1: judge whether not carry out neighbours' vector power of internal layer time distribution, if it is, arbitrarily select a neighbours' vector power not carrying out the distribution of internal layer time, and turns to step 5-2;If it does not, remove current candidate vector power from candidate power vector set U, turn to step 6;

Step 5-2: according to current neighbours vector power, each primary user PU_mCarry out parameter initialization, and be broadcasted to all of from user; Described carry out initialized parameter include send power of information P_m, timing parameter τ, price step-length δ, reservation priceTiming parameter τ initial value is set to 0;

Step 5-3: from user SU_nCalculate relative to each primary user PU_mTransmission rate R_mn:

$R_{mn}=W\&CenterDot;log(1+{\mathrm{\&Gamma;}}_{mn}(1)+{\mathrm{\&Gamma;}}_{mn}(2\left)\right);{\mathrm{\&Gamma;}}_{mn}\left(1\right)=\frac{Q_n\&CenterDot;h_{ss}^{nm}}{N_0};{\mathrm{\&Gamma;}}_{mn}\left(2\right)=\frac{P_m\&CenterDot;h_{ps}^{nm}}{N_0}---\left(12\right)$

Wherein, Γ_mnAnd Γ (1)_mn(2) represent respectively from user SU_nTransmitting terminal ST_nWith primary user PU_mIt send information to from user SU_nReceiving terminal SR_nSignal to noise ratio,WithRepresent respectively from user SU_nTransmitting terminal ST_nArrive from user SU_nReceiving terminal SR_nWith primary user PU_mArrive from user SU_nReceiving terminal SR_nChannel gain, W represents bandwidth, Q_nRepresent from user SU_nThe power of transmission information, P_mRepresent primary user PU_mThe power of forwarding information, N₀Represent noise variance;

Step 5-4: from user SU_nCalculate relative to each primary user PU_mValue of utility u_mn:

$u_{mn}={(t_{mn}\&CenterDot;R_{mn})}^b-c_m^{\mathrm{\&tau;}}\&CenterDot;t_{mn}---\left(13\right)$

Wherein b is the satisfaction factor from user, represents the speed impact on value of utility and 0 < b < 1.

Step 5-5: from user SU_nCalculate relative to each primary user PU_mOptimal timeWith corresponding value of utility

$\frac{\&part;u_{mn}}{\&part;t_{mn}}=0\&DoubleRightArrow;t_{mn}\left(c_m^{\mathrm{\&tau;}}\right)={(b\&CenterDot;{R_{mn}}^b)}^{1/(1-b)}\&CenterDot;{\left(c_m^{\mathrm{\&tau;}}\right)}^{1/(b-1)}---\left(14\right)$

$u_{mn}\left(c_m^{\mathrm{\&tau;}}\right)=(b^{b/(1-b)}-b^{1/(1-b)})\&CenterDot;{R_{mn}}^{b/(1-b)}\&CenterDot;{\left(c_m^{\mathrm{\&tau;}}\right)}^{b/(b-1)}---\left(15\right)$

Step 5-6: from user SU_nDetermining the primary user making its value of utility maximum from M primary user, its numbering is designated as m_n:

$m_n=\arg \underset{x}{max}{u}_{mn}\left(c_m^{\mathrm{\&tau;}}\right)---\left(16\right)$

Step 5-7: from user SU_nDetermine and primary userCorresponding optimum request timeAnd optimum request timeFeed back to primary user

Step 5-8: each primary user PU_mThe all optimum request times from user's request received are sued for peace:

$T_m^{\mathrm{\&tau;}}\left(c_m^{\mathrm{\&tau;}}\right)=\underset{n=1}{\overset{N}{\&Sigma;}}{t}_{mn}\left(c_m^{\mathrm{\&tau;}}\right)---\left(17\right)$

Step 5-9: judge each primary user PU_mWhether all satisfiedIf it is, turn to step 5-11; Otherwise, step 5-10 is turned to;

Step 5-10: timing parameter changes into τ=τ+1; All satisfiedPrimary user PU_mRetention time price is constant, namelyAll it is unsatisfactory forPrimary user PU_mRenewal time priceAnd by the time price c after renewal_mIt is published to N number of from user; Turn to step 5-4;

Step 5-11: timing parameter during note End of Auction is τ=L, according to each from user SU_nRequest timeCalculate primary user's overall utility:

$u=\underset{m}{\&Sigma;}{u}_m=\underset{m}{\&Sigma;}\underset{n}{\&Sigma;}{t}_{mn}^L\&CenterDot;c_m^L---\left(18\right)$

Step 5-12: judge that whether the primary user overall utility value u of current neighbours vector power is more than current primary user overall utility maximum u_max, if it is, update primary user overall utility maximum u_maxFor the primary user overall utility value u of current neighbours vector power, and current neighbours vector power is recorded as candidate power vector, adds candidate power vector set U end to, be subsequently diverted to step 5-1; If it is not, then be not updated, directly to step 5-1.

The beneficial effects of the present invention is: original assignment problem is decomposed into the distribution of outer layer power and two subproblems of internal layer time distribution by the present invention, had both been effectively improved the availability of frequency spectrum, and had achieved again the maximization of primary user's overall utility.

Accompanying drawing explanation

Fig. 1 is the flow chart of the present invention.

Fig. 2 is the cognitive radio networks cooperation transmission first stage schematic diagram of the present invention;

Fig. 3 is the cognitive radio networks cooperation transmission second stage schematic diagram of the present invention;

Fig. 4 is the price convergence graph of the internal layer time distribution primary user of the present invention;

Primary user overall utility figure when Fig. 5 is the outer layer power distribution of the present invention.

Detailed description of the invention

Embodiment 1:

The power of a kind of cognitive radio networks collaboration communication and temporal joint distribution method, described cognitive radio networks includes base station PBS, a M primary user PU_m, m=1LM and N number of from user SU_n, n=1LN;M primary user communicates with base station, forms uplink network; Primary user PU_mBusy channel m, its through-put power is P_m; The acceptable power of base station PBS is P_tot, the vector power P=(P of primary user₁,L,P_M) determined by base station PBS, and meetFrom user SU_nTransmitting terminal be ST_n, receiving terminal is SR_n; Primary user channel is distributed to need collaboration communication from user, each channel can only be assigned at most one within the same time and use from user, each can only assign at most a channel from user; Primary user PU_mThe overall transmission time that can be used to distribute is T_m, primary user PU_mUnit interval price be c_m, from user SU_nObtain the primary user PU of its cooperation transmission_mThe time distributed is t_mn; Primary user PU_mUtility function u_mRepresent, and definition $u_m=\underset{n}{\&Sigma;}{t}_{mn}\&CenterDot;c_m;$

Comprise the following steps:

Step 1: base station PBS initialization power vector P=(P₁,L,P_M), and add in candidate power vector set U;

Step 2: initial internal layer time distribution: setting up distribution model of initial internal layer time and solve and obtain primary user's overall utility, it can be used as the initial value of primary user's overall utility maximum, distribution of described initial internal layer time model is:

$\underset{t}{\max mize}\underset{m}{\&Sigma;}{u}_m=\underset{m}{\&Sigma;}\underset{n}{\&Sigma;}{t}_{mn}\&CenterDot;c_m---\left(1\right)$

It meets constraints:

$\underset{n}{\&Sigma;}{t}_{mn}\&le;T_m,\&ForAll;m---\left(2\right)$

Step 3: take out a candidate power vector from described candidate power vector set as current candidate vector power;

Step 4: outer layer power distributes: generate all neighbours' vector powers of described current candidate vector power;

Step 5: each neighbours' vector power of described current candidate vector power carries out neighbours' internal layer time to be distributed: set up neighbours' internal layer time and distribute model and solve and obtain corresponding primary user's overall utility, updating described candidate power vector set and primary user's overall utility maximum, wherein neighbours' internal layer time distribution model is:

$\underset{t}{\max mize}\underset{m}{\&Sigma;}{u}_m=\underset{m}{\&Sigma;}\underset{n}{\&Sigma;}{t}_{mn}\&CenterDot;c_m---\left(3\right)$

It meets constraints:

$\underset{n}{\&Sigma;}{t}_{mn}\&le;T_m,\&ForAll;m---\left(4\right)$

Step 6: judge whether described candidate power vector set is empty, if it is, turn to step 7; Otherwise, step 3 is turned to;

Step 7: complete power and time distribution, start cooperation transmission: power and time when reaching maximum according to making primary user's overall utility distribute, all primary users distribution is from the request time of user, and during this period of time assists to communicate from user with corresponding power.

Base station PBS initialization power vector P=(P in described step 1₁,L,P_M), and the method adding candidate power vector set U, it is made up of step in detail below:

Step 1-1: set up candidate power vector set U;

Step 1-2: by initialization power vector P=(P₁,L,P_M) add in described candidate power vector set U.

Described step 2 solves the method for distribution of initial internal layer time model, is made up of step in detail below:

Step 2-1: according to described initialization power vector, each primary user PU_mCarry out parameter initialization, and be broadcasted to all of from user; Described carry out initialized parameter include send power of information P_m, timing parameter τ, price step-length δ, reservation priceTiming parameter τ initial value is set to 0;

Step 2-2: from user SU_nCalculate relative to each primary user PU_mTransmission rate R_mn:

$R_{mn}=W\&CenterDot;log(1+{\mathrm{\&Gamma;}}_{mn}(1)+{\mathrm{\&Gamma;}}_{mn}(2\left)\right);{\mathrm{\&Gamma;}}_{mn}\left(1\right)=\frac{Q_n\&CenterDot;h_{ss}^{nm}}{N_0};{\mathrm{\&Gamma;}}_{mn}\left(2\right)=\frac{P_m\&CenterDot;h_{ps}^{nm}}{N_0}---\left(5\right)$

Wherein, Γ_mnAnd Γ (1)_mn(2) represent respectively from user SU_nTransmitting terminal ST_nWith primary user PU_mIt send information to from user SU_nReceiving terminal SR_nSignal to noise ratio,WithRepresent respectively from user SU_nTransmitting terminal ST_nArrive from user SU_nReceiving terminal SR_nWith primary user PU_mArrive from user SU_nReceiving terminal SR_nChannel gain, W represents bandwidth, Q_nRepresent from user SU_nThe power of transmission information, P_mRepresent primary user PU_mThe power of forwarding information, N₀Represent noise variance;

Step 2-3: from user SU_nCalculate relative to each primary user PU_mValue of utility u_mn:

$u_{mn}={(t_{mn}\&CenterDot;R_{mn})}^b-c_m^{\mathrm{\&tau;}}\&CenterDot;t_{mn}---\left(6\right)$

Wherein b is the satisfaction factor from user, represents the speed impact on value of utility and 0 < b < 1;

Step 2-4: from user SU_nCalculate relative to each primary user PU_mOptimal timeWith corresponding value of utility

$\frac{\&part;u_{mn}}{\&part;t_{mn}}=0\&DoubleRightArrow;t_{mn}\left(c_m^{\mathrm{\&tau;}}\right)={(b\&CenterDot;{R_{mn}}^b)}^{1/(1-b)}\&CenterDot;{\left(c_m^{\mathrm{\&tau;}}\right)}^{1/(b-1)}---\left(7\right)$

$u_{mn}\left(c_m^{\mathrm{\&tau;}}\right)=(b^{b/(1-b)}-b^{1/(1-b)})\&CenterDot;{R_{mn}}^{b/(1-b)}\&CenterDot;{\left(c_m^{\mathrm{\&tau;}}\right)}^{b/(b-1)}---\left(8\right)$

Step 2-5: from user SU_nDetermining the primary user making its value of utility maximum from M primary user, its numbering is designated as m_n:

$m_n=\arg \underset{x}{max}{u}_{mn}\left(c_m^{\mathrm{\&tau;}}\right)---\left(9\right)$

Step 2-6: from user SU_nDetermine and primary userCorresponding optimum request timeAnd optimum request timeFeed back to primary user

Step 2-7: each primary user PU_mThe all optimum request times from user's request received are sued for peace:

$T_m^{\mathrm{\&tau;}}\left(c_m^{\mathrm{\&tau;}}\right)=\underset{n=1}{\overset{N}{\&Sigma;}}{t}_{mn}\left(c_m^{\mathrm{\&tau;}}\right)---\left(10\right)$

Step 2-8: judge each primary user PU_mWhether all satisfiedIf it is, turn to step 2-10; Otherwise, step 2-9 is turned to;

Step 2-9: timing parameter changes into τ=τ+1; All satisfiedPrimary user PU_mRetention time price is constant, namelyAll it is unsatisfactory forPrimary user PU_mRenewal time priceAnd by the time price c after renewal_mIt is published to N number of from user; Turn to step 2-3;

Step 2-10: timing parameter during note End of Auction is τ=L, according to each from user SU_nRequest timeCalculate primary user's overall utility, and as primary user overall utility maximum u_maxInitial value:

$u_{max}=\underset{m}{\&Sigma;}{u}_m=\underset{m}{\&Sigma;}\underset{n}{\&Sigma;}{t}_{mn}^L\&CenterDot;c_m^L---\left(11\right)$

Outer layer power distribution in described step 4, the method solving all neighbours' vector powers of current candidate vector power, it is made up of step in detail below:

Step 4-1: definition unit transfer power, is designated as Δ P, and meets Δ P < < P_tot/ M;

Step 4-2: in described current candidate vector power, arbitrarily select two primary users, the power of one of them primary user is reduced or increases unit transfer power Δ P, and the power of another primary user correspondingly increases or reduces unit transfer power Δ P, satisfy condition in the process the power of the power of all primary users and constant and each primary user on the occasion of, namelyP_m> 0; Travel through all situations meeting above-mentioned power transfer process, obtain all neighbours' vector powers of described current candidate vector power.

The method for solving that described neighbours' internal layer time distributes in described step 5 model and the process updating described candidate power vector set and primary user's overall utility maximum are made up of step in detail below:

Step 5-1: judge whether not carry out neighbours' vector power of internal layer time distribution, if it is, arbitrarily select a neighbours' vector power not carrying out the distribution of internal layer time, and turns to step 5-2; If it does not, remove current candidate vector power from candidate power vector set U, turn to step 6;

Step 5-2: according to current neighbours vector power, each primary user PU_mCarry out parameter initialization, and be broadcasted to all of from user; Described carry out initialized parameter include send power of information P_m, timing parameter τ, price step-length δ, reservation priceTiming parameter τ initial value is set to 0;

Step 5-3: from user SU_nCalculate relative to each primary user PU_mTransmission rate R_mn:

$R_{mn}=W\&CenterDot;log(1+{\mathrm{\&Gamma;}}_{mn}(1)+{\mathrm{\&Gamma;}}_{mn}(2\left)\right);{\mathrm{\&Gamma;}}_{mn}\left(1\right)=\frac{Q_n\&CenterDot;h_{ss}^{nm}}{N_0};{\mathrm{\&Gamma;}}_{mn}\left(2\right)=\frac{P_m\&CenterDot;h_{ps}^{nm}}{N_0}---\left(12\right)$

Wherein, Γ_mnAnd Γ (1)_mn(2) represent respectively from user SU_nTransmitting terminal ST_nWith primary user PU_mIt send information to from user SU_nReceiving terminal SR_nSignal to noise ratio,WithRepresent respectively from user SU_nTransmitting terminal ST_nArrive from user SU_nReceiving terminal SR_nWith primary user PU_mArrive from user SU_nReceiving terminal SR_nChannel gain, W represents bandwidth, Q_nRepresent from user SU_nThe power of transmission information, P_mRepresent primary user PU_mThe power of forwarding information, N₀Represent noise variance;

Step 5-4: from user SU_nCalculate relative to each primary user PU_mValue of utility u_mn:

$u_{mn}={(t_{mn}\&CenterDot;R_{mn})}^b-c_m^{\mathrm{\&tau;}}\&CenterDot;t_{mn}---\left(13\right)$

Wherein b is the satisfaction factor from user, represents the speed impact on value of utility and 0 < b < 1.

Step 5-5: from user SU_nCalculate relative to each primary user PU_mOptimal timeWith corresponding value of utility

$\frac{\&part;u_{mn}}{\&part;t_{mn}}=0\&DoubleRightArrow;t_{mn}\left(c_m^{\mathrm{\&tau;}}\right)={(b\&CenterDot;{R_{mn}}^b)}^{1/(1-b)}\&CenterDot;{\left(c_m^{\mathrm{\&tau;}}\right)}^{1/(b-1)}---\left(14\right)$

$u_{mn}\left(c_m^{\mathrm{\&tau;}}\right)=(b^{b/(1-b)}-b^{1/(1-b)})\&CenterDot;{R_{mn}}^{b/(1-b)}\&CenterDot;{\left(c_m^{\mathrm{\&tau;}}\right)}^{b/(b-1)}---\left(15\right)$

Step 5-6: from user SU_nDetermining the primary user making its value of utility maximum from M primary user, its numbering is designated as m_n:

$m_n=\arg \underset{x}{max}{u}_{mn}\left(c_m^{\mathrm{\&tau;}}\right)---\left(16\right)$

Step 5-7: from user SU_nDetermine and primary userCorresponding optimum request timeAnd optimum request timeFeed back to primary user

Step 5-8: each primary user PU_mThe all optimum request times from user's request received are sued for peace:

$T_m^{\mathrm{\&tau;}}\left(c_m^{\mathrm{\&tau;}}\right)=\underset{n=1}{\overset{N}{\&Sigma;}}{t}_{mn}\left(c_m^{\mathrm{\&tau;}}\right)---\left(17\right)$

Step 5-9: judge each primary user PU_mWhether all satisfiedIf it is, turn to step 5-11; Otherwise, step 5-10 is turned to;

Step 5-10: timing parameter changes into τ=τ+1; All satisfiedPrimary user PU_mRetention time price is constant, namelyAll it is unsatisfactory forPrimary user PU_mRenewal time priceAnd by the time price c after renewal_mIt is published to N number of from user; Turn to step 5-4;

Step 5-11: timing parameter during note End of Auction is τ=L, according to each from user SU_nRequest timeCalculate primary user's overall utility:

$u=\underset{m}{\&Sigma;}{u}_m=\underset{m}{\&Sigma;}\underset{n}{\&Sigma;}{t}_{mn}^L\&CenterDot;c_m^L---\left(18\right)$

Step 5-12: judge that whether the primary user overall utility value u of current neighbours vector power is more than current primary user overall utility maximum u_max, if it is, update primary user overall utility maximum u_maxFor the primary user overall utility value u of current neighbours vector power, and current neighbours vector power is recorded as candidate power vector, adds candidate power vector set U end to, be subsequently diverted to step 5-1; If it is not, then be not updated, directly to step 5-1.

With a primary user PU in primary user's network_m, from user network two from user SU_n, SU_kFor example analysis. At primary user PU_mDistribute to SU_nTime t_mnIn, SU_nAt PU_mChannel information, and utilize primary user PU_mCooperation transmission is carried out as relaying. Whole transmission is divided into two stages: the first stage, as in figure 2 it is shown, from user SU_nTransmitting terminal ST_nSend information to its receiving terminal SR_nWith primary user PU_m; Second stage, as it is shown on figure 3, primary user PU_mThe information received decoding is transmitted to SU_nReceiving terminal SR_n。

Fig. 4 gives the through-put power based on each given primary user, the price convergence graph of each primary user when the internal layer time distributes. First respectively select to make its maximum primary user that is benefited according to optimum utility value from user, and request time is submitted to corresponding primary user. Then each primary user calculates time request amount, when request amount is more than its assignable time, can improve time price; When request amount is not more than its assignable time, time price maintenance is constant. When the price adjustment of primary user, selection strategy (including its primary user expecting cooperation and the time to buy) can be changed from user, thus primary user can be affected again can distribute the supply-demand relationship of time. Through 160 iteration adjustments, the time price of all primary users converges to fixed value.

Fig. 5 give base station PBS by the method for Local Search carry out outer layer power distribution time, the change of primary user's overall utility value. From experimental result it can be seen that to all of unit transfer power Δ P, primary user's overall utility value monotone increasing, and unit transfer power Δ P is more little, and the primary user's overall utility maximum obtained is more big.This illustrates when Δ P is less, and having more candidate neighbor vector can searched arrive, and can search bigger value of utility; When Δ P is bigger, neighbours' vector of search reduces, and the vector power that much can produce bigger value of utility is left in the basket. Therefore, unit transfer power is more little, and the primary user's overall utility value searched is more big.

Claims (5)

1. the power of cognitive radio networks collaboration communication and a temporal joint distribution method, described cognitive radio networks includes base station PBS, a M primary user PU_m, m=1LM and N number of from user SU_n, n=1LN; M primary user communicates with base station PBS, forms uplink network; Primary user PU_mBusy channel m, its through-put power is P_m; The acceptable power of base station PBS is P_tot, the vector power P=(P of primary user₁,L,P_M) determined by base station PBS, and meetFrom user SU_nTransmitting terminal be ST_n, receiving terminal is SR_n; Primary user channel is distributed to need collaboration communication from user, each channel can only be assigned at most one within the same time and use from user, each can only assign at most a channel from user; Primary user PU_mThe overall transmission time that can be used to distribute is T_m, primary user PU_mUnit interval price be c_m, from user SU_nObtain the primary user PU of its cooperation transmission_mThe time distributed is t_mn; Primary user PU_mUtility function u_mRepresent, and definition

It is characterized in that: comprise the following steps:

Step 1: base station PBS initialization power vector P=(P₁,L,P_M), and add in candidate power vector set U;

Step 2: initial internal layer time distribution: setting up distribution model of initial internal layer time and solve and obtain primary user's overall utility, it can be used as the initial value of primary user's overall utility maximum, distribution of described initial internal layer time model is:

$\underset{t}{\max mize}\underset{m}{\&Sigma;}{u}_m=\underset{m}{\&Sigma;}\underset{n}{\&Sigma;}{t}_{mn}\&CenterDot;c_m---\left(1\right)$

It meets constraints:

$\underset{n}{\&Sigma;}{t}_{mn}\&le;T_m,\&ForAll;m---\left(2\right)$

Step 3: take out a candidate power vector from described candidate power vector set as current candidate vector power;

Step 4: outer layer power distributes: generate all neighbours' vector powers of described current candidate vector power;

Step 5: each neighbours' vector power of described current candidate vector power carries out neighbours' internal layer time to be distributed: set up neighbours' internal layer time and distribute model and solve and obtain corresponding primary user's overall utility, updating described candidate power vector set and primary user's overall utility maximum, wherein neighbours' internal layer time distribution model is:

$\underset{t}{\max mize}\underset{m}{\&Sigma;}{u}_m=\underset{m}{\&Sigma;}\underset{n}{\&Sigma;}{t}_{mn}\&CenterDot;c_m---\left(3\right)$

It meets constraints:

$\underset{n}{\&Sigma;}{t}_{mn}\&le;T_m,\&ForAll;m---\left(4\right)$

Step 6: judge whether described candidate power vector set is empty, if it is, turn to step 7; Otherwise, step 3 is turned to;

Step 7: complete power and time distribution, start cooperation transmission: power and time when reaching maximum according to making primary user's overall utility distribute, all primary users distribution is from the request time of user, and during this period of time assists to communicate from user with corresponding power.

2. the power of cognitive radio networks collaboration communication according to claim 1 and temporal joint distribution method, it is characterised in that: base station PBS initialization power vector P=(P in described step 1₁,L,P_M), and the method adding candidate power vector set U, it is made up of step in detail below:

Step 1-1: set up candidate power vector set U;

Step 1-2: by initialization power vector P=(P₁,L,P_M) add in described candidate power vector set U.

3. the power of cognitive radio networks collaboration communication according to claim 1 and temporal joint distribution method, it is characterised in that: described step 2 solves the method for distribution of initial internal layer time model, is made up of step in detail below:

Step 2-1: according to described initialization power vector, each primary user PU_mCarry out parameter initialization, and be broadcasted to all of from user;Described carry out initialized parameter include send power of information P_m, timing parameter τ, price step-length δ, reservation priceTiming parameter τ initial value is set to 0;

Step 2-2: from user SU_nCalculate relative to each primary user PU_mTransmission rate R_mn:

$R_{mn}=W\&CenterDot;log(1+{\mathrm{\&Gamma;}}_{mn}(1)+{\mathrm{\&Gamma;}}_{mn}(2\left)\right);{\mathrm{\&Gamma;}}_{mn}\left(1\right)=\frac{Q_n\&CenterDot;h_{ss}^{nm}}{N_0};{\mathrm{\&Gamma;}}_{mn}\left(2\right)=\frac{P_m\&CenterDot;h_{ps}^{nm}}{N_0}---\left(5\right)$

Wherein, Γ_mnAnd Γ (1)_mn(2) represent respectively from user SU_nTransmitting terminal ST_nWith primary user PU_mIt send information to from user SU_nReceiving terminal SR_nSignal to noise ratio,WithRepresent respectively from user SU_nTransmitting terminal ST_nArrive from user SU_nReceiving terminal SR_nWith primary user PU_mArrive from user SU_nReceiving terminal SR_nChannel gain, W represents bandwidth, Q_nRepresent from user SU_nThe power of transmission information, P_mRepresent primary user PU_mThe power of forwarding information, N₀Represent noise variance;

Step 2-3: from user SU_nCalculate relative to each primary user PU_mValue of utility u_mn:

$u_{mn}={(t_{mn}\&CenterDot;R_{mn})}^b-c_m^{\mathrm{\&tau;}}\&CenterDot;t_{mn}---\left(6\right)$

Wherein b is the satisfaction factor from user, represents the speed impact on value of utility and 0 < b < 1;

Step 2-4: from user SU_nCalculate relative to each primary user PU_mOptimal timeWith corresponding value of utility

$\frac{\&part;u_{mn}}{\&part;t_{mn}}=0\&DoubleRightArrow;t_{mn}\left(c_m^{\mathrm{\&tau;}}\right)={(b\&CenterDot;{R_{mn}}^b)}^{1/(1-b)}\&CenterDot;{\left(c_m^{\mathrm{\&tau;}}\right)}^{1/(b-1)}---\left(7\right)$

$u_{mn}\left(c_m^{\mathrm{\&tau;}}\right)=(b^{b/(1-b)}-b^{1/(1-b)})\&CenterDot;{R_{mn}}^{b/(1-b)}\&CenterDot;{\left(c_m^{\mathrm{\&tau;}}\right)}^{b/(b-1)}---\left(8\right)$

Step 2-5: from user SU_nDetermining the primary user making its value of utility maximum from M primary user, its numbering is designated as m_n:

$m_n=\arg \underset{m}{max}{u}_{mn}\left(c_m^{\mathrm{\&tau;}}\right)---\left(9\right)$

Step 2-6: from user SU_nDetermine and primary userCorresponding optimum request timeAnd optimum request timeFeed back to primary user

Step 2-7: each primary user PU_mThe all optimum request times from user's request received are sued for peace:

$T_m^{\mathrm{\&tau;}}\left(c_m^{\mathrm{\&tau;}}\right)=\underset{n=1}{\overset{N}{\&Sigma;}}{t}_{mn}\left(c_m^{\mathrm{\&tau;}}\right)---\left(10\right)$

Step 2-8: judge each primary user PU_mWhether all satisfiedIf it is, turn to step 2-10; Otherwise, step 2-9 is turned to;

Step 2-9: timing parameter changes into τ=τ+1; All satisfiedPrimary user PU_mRetention time price is constant, namelyAll it is unsatisfactory forPrimary user PU_mRenewal time priceAnd by the time price c after renewal_mIt is published to N number of from user; Turn to step 2-3;

Step 2-10: timing parameter during note End of Auction is τ=L, according to each from user SU_nRequest timeCalculate primary user's overall utility, and as primary user overall utility maximum u_maxInitial value:

$u_{max}=\underset{m}{\&Sigma;}{u}_m=\underset{m}{\&Sigma;}\underset{n}{\&Sigma;}{t}_{mn}^L\&CenterDot;c_m^L---\left(11\right).$

4. the power of cognitive radio networks collaboration communication according to claim 1 and temporal joint distribution method, it is characterized in that: the outer layer power distribution in described step 4, the method solving all neighbours' vector powers of current candidate vector power, is made up of step in detail below:

Step 4-1: definition unit transfer power, is designated as Δ P, and meets Δ P < < P_tot/ M;

Step 4-2: in described current candidate vector power, arbitrarily select two primary users, the power of one of them primary user is reduced or increases unit transfer power Δ P, and the power of another primary user correspondingly increases or reduces unit transfer power Δ P, satisfy condition in the process the power of the power of all primary users and constant and each primary user on the occasion of, namelyP_m> 0; Travel through all situations meeting above-mentioned power transfer process, obtain all neighbours' vector powers of described current candidate vector power.

5. the power of cognitive radio networks collaboration communication according to claim 1 and temporal joint distribution method, it is characterised in that: the method for solving that described neighbours' internal layer time distributes in described step 5 model and the process updating described candidate power vector set and primary user's overall utility maximum are made up of step in detail below:

Step 5-1: judge whether not carry out neighbours' vector power of internal layer time distribution, if it is, arbitrarily select a neighbours' vector power not carrying out the distribution of internal layer time, and turns to step 5-2;If it does not, remove current candidate vector power from candidate power vector set U, turn to step 6;

Step 5-2: according to current neighbours vector power, each primary user PU_mCarry out parameter initialization, and be broadcasted to all of from user; Described carry out initialized parameter include send power of information P_m, timing parameter τ, price step-length δ, reservation priceTiming parameter τ initial value is set to 0;

Step 5-3: from user SU_nCalculate relative to each primary user PU_mTransmission rate R_mn:

$R_{mn}=W\&CenterDot;\log \left(1+{\mathrm{\&Gamma;}}_{mn}\left(1\right)+{\mathrm{\&Gamma;}}_{mn}\left(2\right)\right);{\mathrm{\&Gamma;}}_{mn}\left(1\right)=\frac{Q_n\&CenterDot;h_{ss}^{nm}}{N_0};{\mathrm{\&Gamma;}}_{mn}\left(2\right)=\frac{P_m\&CenterDot;h_{ps}^{nm}}{N_0}---\left(12\right)$

Wherein, Γ_mnAnd Γ (1)_mn(2) represent respectively from user SU_nTransmitting terminal ST_nWith primary user PU_mIt send information to from user SU_nReceiving terminal SR_nSignal to noise ratio,WithRepresent respectively from user SU_nTransmitting terminal ST_nArrive from user SU_nReceiving terminal SR_nWith primary user PU_mArrive from user SU_nReceiving terminal SR_nChannel gain, W represents bandwidth, Q_nRepresent from user SU_nThe power of transmission information, P_mRepresent primary user PU_mThe power of forwarding information, N₀Represent noise variance;

Step 5-4: from user SU_nCalculate relative to each primary user PU_mValue of utility u_mn:

$u_{mn}={(t_{mn}\&CenterDot;R_{mn})}^b-c_m^{\mathrm{\&tau;}}\&CenterDot;t_{mn}---\left(13\right)$

Wherein b is the satisfaction factor from user, represents the speed impact on value of utility and 0 < b < 1;

Step 5-5: from user SU_nCalculate relative to each primary user PU_mOptimal timeWith corresponding value of utility

$\frac{\&part;u_{mn}}{\&part;t_{mn}}=0\&DoubleRightArrow;t_{mn}\left(c_m^{\mathrm{\&tau;}}\right)={(b\&CenterDot;{R_{mn}}^b)}^{1/(1-b)}\&CenterDot;{\left(c_m^{\mathrm{\&tau;}}\right)}^{1/(b-1)}---\left(14\right)$

$u_{mn}\left(c_m^{\mathrm{\&tau;}}\right)=(b^{b/(1-b)}-b^{1/(1-b)})\&CenterDot;{R_{mn}}^{b/\left(1-b\right)}\&CenterDot;{\left(c_m^{\mathrm{\&tau;}}\right)}^{b/(b-1)}---\left(15\right)$

Step 5-6: from user SU_nDetermining the primary user making its value of utility maximum from M primary user, its numbering is designated as m_n:

$m_n=\arg \underset{m}{max}{u}_{mn}\left(c_m^{\mathrm{\&tau;}}\right)---\left(16\right)$

Step 5-7: from user SU_nDetermine and primary userCorresponding optimum request timeAnd optimum request timeFeed back to primary user

Step 5-8: each primary user PU_mThe all optimum request times from user's request received are sued for peace:

$T_m^{\mathrm{\&tau;}}\left(c_m^{\mathrm{\&tau;}}\right)=\underset{n=1}{\overset{N}{\&Sigma;}}{t}_{mn}\left(c_m^{\mathrm{\&tau;}}\right)---\left(17\right)$

Step 5-9: judge each primary user PU_mWhether all satisfiedIf it is, turn to step 5-11; Otherwise, step 5-10 is turned to;

Step 5-10: timing parameter changes into τ=τ+1; All satisfiedPrimary user PU_mRetention time price is constant, namelyAll it is unsatisfactory forPrimary user PU_mRenewal time priceAnd by the time price c after renewal_mIt is published to N number of from user; Turn to step 5-4;

Step 5-11: timing parameter during note End of Auction is τ=L, according to each from user SU_nRequest timeCalculate primary user's overall utility:

$u=\underset{m}{\&Sigma;}{u}_m=\underset{m}{\&Sigma;}\underset{n}{\&Sigma;}{t}_{mn}^L\&CenterDot;c_m^L---\left(18\right)$

Step 5-12: judge that whether the primary user overall utility value u of current neighbours vector power is more than current primary user overall utility maximum u_max, if it is, update primary user overall utility maximum u_maxFor the primary user overall utility value u of current neighbours vector power, and current neighbours vector power is recorded as candidate power vector, adds candidate power vector set U end to, be subsequently diverted to step 5-1; If it is not, then be not updated, directly to step 5-1.